Fortinet is expanding its security stack deeper into artificial intelligence with the acquisition of Virtue AI, adding technology designed to continuously test and protect AI models, applications and autonomous agents.
The deal brings Virtue AI’s runtime protection, automated AI validation and agent security technology into the Fortinet AI-Native Security Fabric.
Financial terms were not disclosed. Fortinet said the consideration was immaterial to its business.
The Attack Surface Is Getting Bigger
Traditional enterprise security was built around users, endpoints, applications, networks and cloud infrastructure.
AI introduces an entirely different set of potential targets.
Prompts, models, autonomous agents, Model Context Protocol tools and API calls can all become part of the attack surface.
Fortinet has already started addressing those risks with FortiAIGate, which is designed to protect large language models against threats including prompt injection, data leakage, model poisoning and excessive resource consumption.
Virtue AI takes that strategy further into the AI development and runtime lifecycle.
Red Teaming AI Agents
One of the more interesting capabilities coming from Virtue AI is automated agentic red teaming.
Its technology can test autonomous agents across more than 50 sandboxed environments and 14 high-stakes domains, including simulated prompt injection and MCP-based attacks.
Virtue AI can also continuously evaluate AI systems following model updates and policy fine-tuning.
That matters because AI security is not necessarily a one-time certification problem.
Changing a model, prompt, tool or policy can potentially change the security characteristics of the overall application.
Finding The AI Nobody Approved
Virtue AI also gives organizations visibility into AI applications and agents operating within their environments.
The technology can identify unsanctioned AI, scan MCP tools and source code, monitor agent behavior and block malicious tool calls before they execute.
Real-time guardrails can additionally enforce policies across text, images, video, audio and AI-generated code.
The combination points toward an emerging security model where enterprises need controls not only around access to AI, but around what autonomous systems are actually allowed to do after access is granted.
A Growing AI Security Market
Fortinet is making the acquisition as spending on securing AI environments is expected to grow considerably.
The company cited Gartner projections estimating the market for securing AI ecosystems and agents could increase from $2.8 billion in 2026 to $16.4 billion by 2030.
That creates an obvious opportunity for established cybersecurity vendors.
If AI agents become another major enterprise computing layer, securing them could eventually become as fundamental as protecting endpoints or cloud workloads.
The Bottom Line
Fortinet is betting that AI security will become part of the core enterprise security stack rather than remain a separate category of specialized tools.
Virtue AI gives it technology for testing AI before deployment, monitoring it in production and stopping dangerous behavior as it happens.
That becomes increasingly important as enterprises move from AI assistants that recommend actions to autonomous agents capable of taking them.
The more authority organizations give AI, the more important it becomes to continuously verify what those systems are actually doing.